Re: VMs: NSU review of Rugg (2003)...

[Gordon Rugg <g.rugg@xxxxxxxxxxxxxx>]

Dear Gabriel,

Gabriel Landini wrote:

> On Thursday 18 December 2003 08:46, Gordon Rugg wrote:
> > padding to surround a meaningful ciphertext. At present, using Occam's
> > razor, the hoax hypothesis looks like the simplest explanation,
>
> I am interested to know what makes you believe so.
> So far the only argument seems "because it cannot be read".

There are two reasons. One is that there are various hypotheses which are
compatible with the table and grille; the main ones are:
1: meaningless gibberish with no meaningful text
2: meaningless gibberish as padding, plus some meaningful text
3: meaningful text

I've identified a couple of ways in which the table and grille could be used for
reversible encoding of material. It's possible to eliminate the first fairly
definitely, because of the degree of repetition in successive words of the
manuscript. The second also appears unlikely to have been used because of the
degree of repetition between successive lines of the manuscript.

So, in relation to the three main hypotheses above, I think we have:
1: possible, using table and grille
2: possible, but no evidence at present for the presence of any meaningful text
3: unlikely with the coding mechanisms I've identified, and no evidence at
present for the use of a different coding mechanism

In brief, I'm simply saying that the manuscript can be explained as a hoax
containing only meaningless gibberish; there's no evidence at present of any
meaningful text hidden among gibberish. That doesn't mean that we can definitely
exclude meaningful text, and I'd be delighted if someone found some.

>
>
> Other Occam's razors are possible.
>
> > I think that a fruitful way forward would be to identify the parts of the
> > VMS which could be produced using tables and grilles, and see what's left;
> > if there is a ciphertext, it would probably be in the remaining bits.
>
> That is interesting too. How complex should this grille have to be to explain
> ALL the vms?
>

I suspect that about seven or eight tables were used; I don't know how many
grilles. That suspicion comes from two directions. From attempting to replicate
the manuscript, I found that the table needs to be changed periodically, and
producing a manuscript the size of the VMS would take about half a dozen tables.
>From the textual evidence in the manuscript, I suspect that the difference
between Voynich A and B reflects table A1 and table B1; other sections of the
manuscript are probably derived from modifications to table A1 and table B1 (e.g.
table A2, table B2, etc).

You can produce some pretty sophisticated restrictions on character distributions
very easily using tables and grilles - for instance, two characters which are
individually equally common, but where one is four times as common before a third
character as the other. It's easy to alter the relative proportions, ranging from
no co-occurrence to total co-occurrence. For entropy levels, Letter Serial
Correlation, etc, you should be able to manipulate values by changing the degree
of regularity in the table layout (e.g. a "qo" in every fourth prefix cell, a
"shek" in every sixth midfix cell, etc). That would also probably be affected by
the way you broke down a word. I've used Stolfi's analysis for convenience, but I
suspect that Nick Pelling and/or Philip Neal are more likely to be correct.

For what it's worth, there are also some things that can be done with tables and
grilles which appear not to have been used in the VMS. Examples include vowel
harmony, a simulacrum of case and tense endings, and phrases. I suspect that some
of these are deliberate omissions (for instance, the presence of apparent case
and tense endings would significantly reduce the number of plausible ciphering
systems which a would-be cracker needed to consider), but that's a guess.

Best wishes,

Gordon


>
> Cheers,
>
> Gabriel
>
> ______________________________________________________________________
> To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
> unsubscribe vms-list

______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list